[MUD-Dev] Morphable worlds, Reset based systems revisited

[John Robert Arras]

On Sun, 27 Oct 2002, Ted L. Chen wrote:
> John Robert Arras wrote
 
>> Could you give me an example of what you mean in terms of placing
>> wilderness areas like swamps or forests into some kind of a world
>> map?
 
> Yup, design on a larger scale.  If I understand you correctly,
> your algorithm would generate several swamp-like tiles next to
> each other to create a swamp?

> So, why resolve to the level of individual swamps tiles if the
> user's view it as just a swamp?  "You are now in Swamp0A.  You are
> now in Swamp0B.  You are now in Swamp1B."  Instead, just say "You
> are now in the Dagobah swamplands."  You're describing the Dagobah
> swamplands anyways in your algorithm (the set of features it
> uses).

Ok, I am already doing this. The code works on a "global" level to
decide where to put areas near each other based on overall terrain
types, then each area is generated and then they're all linked
together. The areas get individual names and each room gets an
individual name, and the process is one of coarse-->refined feature
creation.

>   Something that trumps this of course is if you value continuity
>   of scale.  Abstractions don't work well when you try to visually
>   represent it ;) 

I want some continuity of scale, but I also want locally interesting
things. I do put some restrictions on what can or can't appear
within an area. For example, a patch of snow will not appear in a
swamp, nor will it appear in a desert. But, a patch of grass may
appear anywhere.

> The other option I gave was to go the algorithm route, but design
> the game-mechanics such that there's meaning to the generated
> tiles, instead of them being just decorative.  If you move your
> game into a more strategic war-game genre, then the layout of
> swamps and forests would be very useful.  In this case, you could
> probably get away with some repetitiveness in zones since people
> aren't concentrating at the level of individual zones.

I'm not looking for meaning. What I'm looking for is areas that are
interesting enough that people don't get bored with them. My hope is
that I can create enough different interestingng things that the
players don't get bored. Since I'm using all text and my rooms
aren't coordinate-based, this is much easier. I can cheat quite a
bit. The interest comes in terms of the shapes of the areas and how
useful they are for hunting and ambushes. There are three levels of
connection between areas: land, underground, and catwalks in the
trees. I hope that players will learn to set ambushes and attempt to
take over societies at the chokepoints that get generated.

> But if you stay with the current trends of MOGs, and focus on a
> in-your-shoes perspective, then being stuck in a set of zones that
> look all similar is bland.  From a players' viewpoint, there's no
> care about the fact the swamp I'm in now is different from the
> forest 1 mile away because the game-mechanics concentrate on that
> rock in front of me.

I hope to get it good enough that the areas are different. I might
explain the current algorithm in another note, since it will take a
bit to explain it.

> Hmm.. didn't I pass that rock 10 mins ago?  ;)

Ahh...the power of text. It's not a rock, it's a round moss-covered
granite boulder.

Or perhaps 

  a [<rock-shape>] [<rock-description>] [<rock->type>] <rock-name>

  a round          moss-covered         granite        boulder

or 

  a [<rock-shape>] [<rock-description>] <rock-name> made of [<rock-type>]

You know. The generation thing. Use your imagination to try to trick
the players into thinking that these are all-new rocks, not those
boring rocks they saw a few minutes ago.


John


_______________________________________________
MUD-Dev mailing list
MUD-Dev at kanga.nu
https://www.kanga.nu/lists/listinfo/mud-dev